The mating-type MAT locus of Saccharomyces cerevisiae has proven to be a remarkable system for studying many different aspects of a programmed genetic switch, initiated by a site-specific double- strand break (DSB). The ability to analyze DNA from cells synchronously undergoing recombination has made it possible to dissect many steps in the process of gene conversion, as a model for general homologous recombination. The continuation of this project will focus on two major questions. First, what is the molecular mechanism of gene conversion initiated by the HO endonuclease that leads to MAT switching? Experiments are proposed to more directly analyze the participation of DNA polymerases and other replicaiton factors, as well as many recombination proteins, topoisomerases and helicases, by using chromatin immunoprecipitation techniques. Additional experiments will analyze break-induced replication, to establish its relation to gene conversion and to explore why this process, but not gene conversion, occurs independently of the Rad51 strand exchange protein and several of its accessory factors. A search will be made to identify genes responsible for a Rad1, Msh2-independent pathway for the removal of nonhomologous DNA tails from the ends of DSBs, allowing recombination to be completed. Second, we will analyze the phenomenon of donor preference, the mating type-dependent choice of using the HML or HMR donors during MAT switching. Donor preference is controlled by two mechanisms: the MATalpha-dependent inactivation of the left arm of chromosome III, coupled to the MATa-dependent activation of the entire left arm by the Recombination Enhancer (RE). Further dissection of the RE will be undertaken, to identify the important cis-acting sequences and the trans-acting proteins that bind to these regions. Additional donor preference mutants will also be sought by a genetic screen that takes advantages of two independent criteria. The way that the RE is able to control recombination at long-distance will be analyzed by direct visualization of the in vivo movement of chromosome regions by following the three-dimensional movement of a fluorescent LacI-GFP fusion protein bound to arrays of the LacO sequence inserted at specific chromosome locations. Finally, we will undertake to characterize a cis-acting "cold spot" that is at least partly responsible for the MATalpha-dependent inactivation of the left arm.